Empowering Elected Women Representatives in Local Democratic Institutions

The treatment will provide targeted training to the women elected representatives in the treatment GPS focused on enhancing skills and knowledge related to flood management, public service delivery, and leadership. Nested treatments will also offer training to improve self-efficacy and confidence, as well as communication skills. Topics will include resource allocation, community mobilization, emergency response, and access to public provisions during and after floods.

2024-06-01

2026-06-01

EWRs: awareness, participation in GP decision making, quality of governance, confidence, communication skills, public goods provision, and empowerment.
Community welfare and resilience to floods

EWR
Awareness - quiz about roles and responsibilities and resources available for flood management in GP;
Participation in GP- attendance in meetings, speaking in meetings;
Quality - roles EWRs play, services they monitor, meetings they hold in their ward and with public officials, and petitions they submit;
Confidence -generalized self-efficacy scale;
Public goods provision -mapping public goods
Empowerment- Index of decisions that EWR can make and autonomy over mobility
Community welfare - examining expenditure, food vulnerability, health (Activities of daily life and mental health related survey questions),
Adaptation and resilience- flood preparation and precautionary actions, age of marriage of girls during floods, safety, welfare outcomes listed above after floods

attidtudes towards gender roles of males, and females ERs;
crimes against females; improvement in bargaining power at home for EWRs, allocation of resources to children

Attitudes - creating an index of agreement with statements about progressive norms and survey questions through vignettes
crimes- survey questions about assault, trafficking, domestic violence
bargaining power- how much of the household budget is spent on preferred goods
allocation for health, nutrition, and schooling of children

We are conducting a cluster randomized controlled trial in Darbhanga district, Bihar, India, one of the country’s most flood-prone areas. The study tests whether leadership and capacity-building programs for elected women representatives (EWRs) in Gram Panchayats (village governments) improve governance and flood preparedness.

A total of 105 flood-vulnerable Gram Panchayats were randomly assigned to one of two groups:

Control group (35 GPs) – no intervention.

Treatment group (70 GPs) – in Year 1, all receive a three-day “boot camp” training on their roles and responsibilities, local governance, gender equity, and flood management.

After one year, the treatment group is split into two equal subgroups: one receives additional training in self-efficacy, the other in communication skills. The full intervention spans two years.

We will track changes in leadership capacity, flood preparedness, and local governance through surveys of elected representatives and households at baseline (2024), midline (2025), and endline (2026).

Not available

Randomization done by a computer

Gram panchayats (GPs)

Yes

105

approximately 2205, including elected representatives and 12 additional households

Treatment T0: Exposure and awareness 70 GPS> T1: Training in self-efficacy 35 GPs, T2: training in communications and oral efficacy 35 GPs
Control 35 GPs

Power calculation (MDEs). We computed MDEs for a two-sided test (α=0.05) with 80% power in a cluster-randomized design at the Gram Panchayat (GP) level, accounting for clustering. For the main comparison (any treatment vs control; 70 vs 35 GPs), assuming average cluster sizes of 12 households/GP and 9 elected representatives/GP and ICC=0.05, the MDEs are 0.208 SD (household outcomes) and 0.229 SD (elected-representative outcomes). For the Year-2 sub-arm comparison (self-efficacy vs communication; 35 vs 35 GPs), the MDEs are 0.241 SD (household) and 0.264 SD (elected representatives). For binary outcomes, we report percentage-point equivalents using Sqroot(p(1−p) ): at p=0.50, these correspond to ~10.4–13.2 pp; At p=0.30, ~9.6–12.1 pp. Primary outcomes that are indices are standardized to SD=1 by construction.